The formation stability, hydrolytic behavior, mass spectrometry, DFT study,and luminescence properties of trivalent lanthanide complexes of H2ODO2A†
C. Allen Chang,*a,b,c I-Fan Wang,a Hwa-Yu Lee,b Ching-Ning Meng,b Kuan-Yu Liu,a,b Ya-Fen Chen,d,e
Tsai-Hua Yang,b Yun-Ming Wanga and Yeou-Guang Tsayd,e
Received 6th July 2012, Accepted 11th September 2012DOI: 10.1039/c2dt31479g
The trivalent lanthanide complex formation constants (log Kf ) of the macrocyclic ligand H2ODO2A(4,10-dicarboxymethyl-1-oxa-4,7,10-triazacyclododecane) have been determined by pH titrationtechniques to be in the range 10.84–12.62 which increase with increasing lanthanide atomic number,and are smaller than those of the corresponding H2DO2A (1,7-dicarboxylmethyl-1,4,7,10-tetraazacyclododecane) complexes. The equilibrium formation of the dinuclear hydrolysis species,e.g. Ln2(ODO2A)2(μ-OH)
+ and Ln2(ODO2A)2(μ-OH)2, dominates over the mononuclear species, e.g.LnODO2A(OH) and LnODO2A(OH)2
−. Mass spectrometry confirmed the presence of [Eu(ODO2A)]+,[Eu(ODO2A)(OH)+H]+, [Eu2(ODO2A)2(OH2)2+H]
+, [Eu(ODO2A)(OH)2]− and [Eu2(ODO2A)2(OH2)3]
−
species at pH > 7. Density function theory (DFT) calculated structures of the EuODO2A(H2O)3+ and
EuDO2A(H2O)3+ complexes indicate that three inner-sphere coordinated water molecules are arranged
in a meridional configuration, i.e. the 3 water molecules are on the same plane perpendicular to thatof the basal N3O or N4 atoms. However, luminescence lifetime studies reveal that the EuODO2A+
and TbODO2A+ complexes have 4.1 and 2.9 inner-sphere coordinated water molecules, respectively,indicating that other equilibrium species are also present for the EuODO2A+ complex. The respectiveemission spectral intensities and lifetimes at 615 nm (λex = 395 nm) and 544 nm (λex = 369 nm) of theEuODO2A+ and TbODO2A+ complexes increase with increasing pH, consistent with the formation ofμ-OH-bridged dinuclear species at higher pH. Additional DFT calculations show that each Y(III) ion is8-coordinated in the three possible cis-[Y2(ODO2A)2(μ-OH)(H2O)2]
+, trans-[Y2(ODO2A)2(μ-OH)-(H2O)2]
+ and [Y2(ODO2A)2(μ-OH)2] dinuclear complex structures. The first and the second include6-coordination by the ligand ODO2A2−, one by the bridged μ-OH ion and one by a water molecule.The third includes 6-coordination by the ligand ODO2A2− and two by the bridged μ-OH ions. The twoinner-sphere coordinated water molecules in the cis- and trans-[Y2(ODO2A)2(μ-OH)(H2O)2]
+ dinuclearcomplexes are in a staggered conformation with torsional angles of 82.21° and 148.54°, respectively.
Introduction
Trivalent lanthanide ions (Ln3+) are hard Lewis acids and inaqueous solution they are usually 8–10 coordinated. When theLn3+ ion, particularly Gd3+, is coordinated with multidentate
ligands that allow usually one inner-sphere coordinated watermolecule, the resulting stable complexes have been used aseffective magnetic resonance imaging (MRI) contrast agents,e.g. Magnevist, ProHance and Omniscan.1 If the coordinatedligand or inner-sphere coordinated water molecules on the Ln3+
ion are able to be bio-activated (bio-responsive) or allow anionreplacement, these systems could be used as molecular imagingagents2 or luminescent anion sensors.3 On the other hand, thecoordinated water molecules could be hydrolyzed at relativelylower pH which makes them potentially good candidates asartificial nucleases, peptidases and hydrolases.4
Early research works on the hydrolysis of trivalent lanthanideions are summarized in an excellent monograph.5 Although anumber of ligand-controlled self-assembly of polynuclear lantha-nide-oxo/hydroxo complexes have been recently synthesized andcharacterized by crystallography,6 except for the first hydrolysisconstants leading to the formation of the mononuclear Ln-OH2+
species,7 the equilibrium constants for the formation of†Electronic supplementary information (ESI) available. See DOI:10.1039/c2dt31479g
aDepartment of Biological Science and Technology, National ChiaoTung University, No. 75 Po-Ai Street, Hsinchu, Taiwan 30039, R. O. C.bDepartment of Biomedical Imaging and Radiological Sciences,National Yang-Ming University, No. 155, Sec. 2, Li-Nong St., Beitou,Taipei, Taiwan 112, R. O. C. E-mail: [email protected];Fax: +886-2-28201093; Tel: +886-2-28201091cBiophotonics & Molecular Imaging Research Center (BMIRC),National Yang-Ming University, Taipei, Taiwan, R. O. C.dProteomics Research Center, National Yang-Ming University, Taipei,Taiwan, R. O. C.eInstitute of Biochemistry & Molecular Biology, National Yang-MingUniversity, Taipei, Taiwan, R. O. C.
6+ could only be determined in high ionic strengthmedia with limited accuracy, due to the complexity of varioushydrolysis equilibria at high pH and easy hydroxideprecipitation.
Formation of trivalent lanthanide complexes with multidentateligands tends to reduce the number of the inner-coordinatedwater molecules and allow possibly better control of theirhydrolysis behaviors. For example, the anionic EuEDTA−
(EDTA4− = ethylenediamine-N,N,N′,N′-tetraacetate ion) andneutral EuHEDTA (HEDTA3− = N-hydroxylethyl(ethylene-diamine)-N,N′,N′-triacetate ion) complexes have the respectivecoordinated water hydrolysis constants of pKh = 12.488a andpKh = 10.1.8b However, for applications as artificial nucleases,recent studies suggested the use of stable and positively chargeddinuclear lanthanide complexes with at least two or more inner-sphere coordinated water molecules.9 In our laboratory, we havefound that the EuDO2A+ complex with 3 inner-sphere coordi-nated water molecules was able to promote phosphodiester bondhydrolysis with appreciable rates at pH 9–10 (DO2A2− is thedeprotonated form of H2DO2A, i.e. 1,7-dicarboxylmethyl-1,4,7,10-tetraazacyclododecane, Scheme 1).10 It was proposedthat the mono-hydroxo-bridged Eu2(DO2A)2(μ-OH)(OH)(H2O)3and the di-hydroxo-bridged Eu2(DO2A)2(μ-OH)2(H2O)2 specieswere more reactive than the mononuclear species. However, dueto the slow complex formation rates, the formation constants forthese μ-OH bridged dinuclear species could not be determineddirectly and accurately to further confirm exactly which specieswas the more reactive.11
Previously, we have found that the Ln(III) complex formationconstants12 of a number of oxaaza-macrocyclic diacetic acidligands, i.e. H2K21DA (dapda, 1,7-diaza-4,10,13-trioxacyclo-pentadecane-N,N′-diacetic acid,12b Scheme 1) and H2K22DA(dacda, 1,10-diaza-4,7,13,16-tetraoxacyclooctadecane-N,N′-di-acetic acid,12a Scheme 1) could be conveniently determined bypH titration techniques. In order to understand the formationstability, hydrolysis behaviors, and structures of trivalentlanthanide complexes better, we have synthesized the oxaaza-macrocyclic ligand, 4,10-dicarboxymethyl-1-oxa-4,7,10-triaza-cyclododecane (H2ODO2A, Scheme 1),13 and have determinedtheir trivalent lanthanide complex stability constants and variousformation constants involving hydroxide species. We have alsostudied mass spectrometry, solution luminescence properties14
and density function theoretical (DFT) predictions of their struc-tures.15 The results are reported in this paper.
Results and discussion
Synthesis and purification of H2ODO2A
The synthesis of the ligand H2ODO2A has been reported pre-viously and is rather straightforward.13 Protection of the middlesecondary ring nitrogen atom from carboxymethylation can bedone by controlling reaction solution pH because protonation ofthis more basic nitrogen atom occurs at high pH, i.e. the logarith-mic first protonation constant of 1-oxa-4,7,10-triazacyclo-dodecane is 10.11.16a Although the previously reported synthesisfor the carboxymethylation reaction was performed at pH 8.5, inour hands we found that by keeping the pH to ca. 8.0 and usingtwo equivalents of bromoacetate, the tris-carboxy-methylationby-product could be greatly reduced. This final purification ofH2ODO2A was performed by first passing the last reactionproduct solution through an anion exchange column, concentrat-ing it, and then recrystallizing it in ethanolic HCl solution.
Ligand protonation constants and protonation sites ofH2ODO2A
The logarithmic protonation constants of H2ODO2A determinedby the potentiometric method are 11.08 ± 0.02, 5.96 ± 0.04,2.85 ± 0.10, and 1.94 ± 0.10. These values are a little lower thanthose reported previously,13 i.e. 11.24, 6.02, 2.94. The previouslyreported NMR titration experiments at various pH solutions werealso repeated by us and similar results were obtained (Fig. S1,ESI†). The data concerning the protonation sequence were bestexplained by first protonation at the secondary ring nitrogenand second protonation on the tertiary ring nitrogen atomswith simultaneous partial deprotonation of the secondary ringnitrogen atom.
Stabilities of LnODO2A+ complexes
Fluorescence titration by the molar ratio method indicated thatEu3+ and ODO2A2− at pH 6.7 form a 1 : 1 complex (λex =395 nm, λem = 615 nm; data not shown). Unlike those ofDO2A2− complexes,11 the kinetics are relatively faster for theformation reactions between the trivalent lanthanide ions and theligand ODO2A2− due to faster equilibrium pH establishmentafter each addition of standard base, and the stability constantscould be conveniently determined by potentiometric pH titrationtechniques. The logarithmic formation constants of theLnODO2A+ complexes (including YODO2A+) and theirhydrolysis species (vide infra) are listed in Table 1. Scheme S1†shows the definitions of formation reactions and constants forthese hydrolysis species (ESI,† L = ODO2A). For comparisonpurposes, the stability data of Ln3+ complexes of three structuralanalogues of ODO2A2−, i.e. DO2A2−, K21DA2− and K22DA2−
are listed in Table S1 (ESI†) and plotted in Fig. 1. Fig. S2(ESI†) shows some selected pH titration curves of LnODO2A+.
The formation constants of the LnODO2A+ complexes are inthe range log Kf 10.84–12.62 and increase with increasing
Scheme 1 Structural formulas of H2DO2A (upper-left), H2ODO2A(upper-right), H2K21DA (lower-left) and H2K22DA (lower-right).
lanthanide atomic number, to a greater extent for the lighterLn(III) ions (i.e. La–Eu) and a lesser extent for the heavier Ln(III)ions (i.e. Gd–Lu). This indicates that electrostatic interaction isdominant between Ln3+ and ODO2A2−. The phenomenon ofgadolinium break seems also apparent. The LnODO2A+ stabilityis usually smaller (i.e. up to 1.27log K units) than that of thecorresponding LnDO2A+. Thus, substituting one secondarynitrogen donor atom of the cyclen macrocycle of DO2A2− by anether donor atom to result in the ligand ODO2A2− decreases theoverall Ln3+ formation stability. At least three reasons are appa-rent. The first is that the basicity of ODO2A2− as represented bythe summation of the first three protonation constants (Σlog K =19.89) is less than that of DO2A2− (Σlog K = 24.34). Second,the rigidity of the macrocycle of ODO2A2− is relatively less thanthat of DO2A2− as manifested by the relatively fasterLnODO2A+ formation and dissociation kinetics (to be reportedelsewhere). Third, the LnODO2A+ complex is less symmetricalthan the corresponding LnDO2A+ and is more distorted andresults in lower stabilization energy.
For macrocyclic ligands with ionizable coordinating pendantfunctional groups, the electrostatic interactions between theLn(III) ions and the ligand are expected to increase as the chargedensity of the Ln(III) ion increases across the series due to lantha-nide contraction. On the other hand, the match between the sizeof the macrocyclic cavity and the Ln(III) ion could tune thecomplex formation selectivity. A more careful examination ofthe stability trend of the lighter Ln(III)-ODO2A2− complexes (i.e.La–Eu) indicates that it is similar to those of LnDO2A+ andLnK21DA+, as well as other macrocyclic aminopolycarboxylatecomplexes of DOTA4− (1,4,7,10-tetra-azacyclododecane-1,4,7,10-tetraacetate ion), TETA4− (1,4,8,11-tetraazacyclotetra-decane-1,4,8,11-tetraacetate ion), PEPA5− (1,4,7,10,13-pentaaza-cyclo-pentadecane-N,N′,N′′,N′′′,N′′′′-pentaacetate ion) and HEHA6−
(1,4,7,10,13,16-hexaaza-cyclooctadecane-N,N′,N′′,N′′′,N′′′′,N′′′′′-hexaacetate ion),17 consistent with the notion that electrostaticinteraction dominates the lighter Ln(III)-cyclen macrocyclic
complex formation. The ODO2A2− macrocycle is too small toexert a size effect towards the lighter and larger Ln(III) ions.However, for the heavier Ln(III)-ODO2A2− complexes (i.e.Gd–Lu), the observed trend suggests factors such as better fitbetween metal ion radius and ligand cavity size may be impor-tant. As the number of the ether oxygen atoms is increased andthe macrocycle ring becomes larger, the flexibility of the macro-cycle ring is increased and so are the modulations of the sizeeffect to Ln(III) ion selectivities. This probably could be shownby the Δlog Kf,range (i.e. log Kf,highest − log Kf,lowerest) values.Among the ligands illustrated in this paper, the Δlog Kf,range
values are greater for lanthanide complexes with dominant ionicinteractions (e.g. the respective Δlog Kf,range values for DO2A
2−,TETA4−, PEPA5−, and HEHA6− are 2.37, 2.57, 3.14 and 5.16with selectivities towards the heavier lanthanides) and smaller
Table 1 Formation constants of the Ln(III)-ODO2A2− complex systems, 25 °C, ionic strength 0.1a,b
a The overall protonation constants for ODO2A2− are: log β1, 11.08; log β2, 17.04; log β3, 19.89; log β4, 21.83.b The formation constants were the
averaged values of three to five determinations. Values in parentheses are the standard deviations of each formation constant. The averaged standarddeviation values are as follows: log Kf, 0.01; log β111, 0.16; log β22-1, 0.13; log β22-2, 0.14; log β22-3, 0.20; log β22-4, 0.17. The relatively largerstandard deviation values for the species involving hydroxide formation (i.e. Ln2L2H-3 and Ln2L2H-4) are due to slower equilibria as compared tothose of LnL.
Fig. 1 Plots of the log Kf values for the Ln(III) complexes of ODO2A2−
(△), DO2A2− (●), K21DA2− (○) and K22DA2− (▼). Data are fromTable S1.†
for those with more cavity size modulations (e.g. the respectiveΔlog Kf,range values for ODO2A
2−, K21DA2− and K22DA2− are1.78, 1.74, and 1.39 with selectivities varied from that forheavier, middle and lighter lanthanides, respectively). Note thatan unprecedented selectivity for lighter lanthanides has beenreported for a new macrocyclic ligand, N,N′-bis[(6-carboxy-2-pyridyl)methyl]-4,13-diaza-18-crown-6 with a Δlog Kf,range of6.86.18 For analogous cationic complexes of LnODO2A+,the stability trend is roughly LnK22DA+ > LnDO2A+ ≥LnODO2A+ > LnK21DA+ for the lighter La–Eu complexes,with a few exceptions of the LnK22DA+ and LnDO2A+ com-plexes (i.e. Ln = Nd and Eu). For the heavier Gd–Lu complexes,the trend is roughly LnDO2A+ > LnODO2A+ > LnK22DA+ >LnK21DA+ presumably due to the larger ring size flexibilities ofthe K21DA/K22DA ligands and the interplay of the overallgreater basicity and ring rigidity of the DO2A/ODO2A ligands.
The protonation constants of the lanthanide complexes (log K′)could be obtained by the differences between the correspondinglog β111 and log Kf(log β110) values. Our titration data allowthe calculations of these values only for six lighter LnODO2A+
(Ln = La–Eu) complexes and they are in the range 3.28–3.98.These values are similar to those of the LnDO2A+ complex, i.e.3.20–3.9819a and the LnK21DA+ complex, i.e. 3.47–4.2419b
determined from kinetic studies. These values are consistentwith the proposition that protonation is on the carboxylate func-tional groups of these complexes.
Hydrolytic behaviors of the LnODO2A+ complexes in aqueoussolution
It is also noted that the titration curves became much morecomplex after roughly 5 equivalents of base were added(Fig. S2, ESI†). Fitting these data to the equilibrium andhydrolysis model shown in Fig. 2 leads to much more interestingand significant results involving μ-OH bridged dinuclearLn2L2(μ-OH)n (n = 1–2) species formation with smaller devia-tions. Note that when mononuclear hydroxide species wereincluded, the data fittings in most cases either wouldn’t convergeor converged with larger deviations. To our knowledge, althoughseemingly straightforward, this general behavior has not beenreported previously in a complete and systematic fashion. It isalso noted that this scheme was not reported when transitionmetal (e.g. Cu(II) and Zn(II)) oxatriaza complexes were studiedbecause those complexes preferred simple stepwise mononuclearhydroxide species formation.16
Depending on the ionic radius of the Ln3+ ion, the number ofinner-sphere coordinated water molecules could be 2, 3, or 4 ifthe 6-coordinating ODO2A2− is already coordinated to the Ln3+
ion. For a smaller Ln(III) ion (e.g. Yb3+ and Lu3+) and for themacrocyclic ligands which result in the formation of Ln(III) com-plexes with two inner-sphere coordinated water molecules (e.g.DO2A2−, ODO2A2− and Ln-K21DA2−,14), the number of majordinuclear μ-OH bridged species would be two, i.e. Ln2L2(μ-OH)and Ln2L2(μ-OH)2. More careful examinations of the speciationdiagrams (e.g. Fig. 3) reveal that: (1) the μ-OH bridgedLn2L2(μ-OH) species is always formed before LnL(OH); (2)Ln2L2(μ-OH) is formed prior to Ln2L2(μ-OH)2, Ln2L2(μ-OH)2(OH)
−, and Ln2L2(μ-OH)2(OH)22−, and its stability gradu-
ally increases slightly with increasing atomic number, however,the relative species abundances are not always easily predicable;(3) the equilibrium formation of the mononuclear species (e.g.LnL(OH) and LnL(OH)2
−) is less significant as compared to thedinuclear species; (4) the log β22-3 and log β22-4 values are sub-jected to larger uncertainties due to relatively larger measurederrors at higher pH, and systematic trends are not obviouslyobserved.
Fig. 3 shows the speciation diagrams of the La(III)-ODO2A2−,Eu(III)-ODO2A2− and Yb(III)-ODO2A2− complex systems. It isobserved that although maximum EuODO2A+ formation occursat a lower pH (i.e. pH 5.5) than that of LaODO2A+ (pH ∼ 6.0),the maximum formations of the μ-OH bridged dinuclear La2L2-(μ-OH)n(OH)m (m,n = 1,2) species all occur at lower pH than thecorresponding Eu2L2(μ-OH)n(OH)m (m,n = 1,2) species. On theother hand, 95% of YbODO2A+ forms at pH 5.4 which is lowerthan those of LaODO2A+ and EuODO2A+, the maximum for-mations of the μ-OH bridged dinuclear Yb2L2(μ-OH)n(OH)m(m,n = 1,2) species both occur at lower pH than the corres-ponding La2L2(μ-OH)n(OH)m (m, n = 1,2) species. Similar datahave also been found for other macrocyclic ligand systems suchas H2NO2A (4,7-dicarboxymethyl-1,4,7-triazacyclononane) andH2ONO2A (4,7-dicarboxymethyl-1-oxa-4,7-diazacyclononane)but could not be obtained with traditional linear ligands such asethylenediamine-N,N′-diacetic acid (H2EDDA) nor those witheasy hydroxide precipitation. These results will be reportedelsewhere.
Mass spectrometry
Both the positive and negative ESI-MS spectral data of theEu(III)-ODO2A2− complex system confirmed the presence of thehydrolytic, mono- and di-nuclear species. The ESI-MS(+) peaksfor the [Eu(ODO2A)]+ species (m/z at 438, 440), [Eu(ODO2A)-(OH) + H]+ species (m/z at 456 and 458) and for the [Eu2-(ODO2A)2(OH2)2 + H]+ species (m/z at 911, 913 and 915)could be observed at pH 7.36 (Fig. 4). The ESI-MS(−) peaks forthe [Eu(ODO2A)(OH)2]
− species (m/z at 472 and 474) and forthe [Eu2(ODO2A)2(OH2)3]
− species (m/z at 927, 929 and 931)could be observed at pH 8.28 (Fig. S3, ESI†). It is noteworthythat, due to the distinct ionization energies of the mononuclearand dinuclear species, it is very likely that these ions in solutioncould not be quantitatively compared according to their signalsin mass spectrometry.
Fig. 2 A proposed general hydrolysis scheme for some of trivalentlanthanide ODO2A2− complexes.
The density function theory (DFT) calculated lowest energystructures of the EuDO2A(H2O)3
+ and EuODO2A(H2O)3+ com-
plexes are shown in Fig. 5. The DFT calculation method usedwas similar to those published and established for manycomplex systems including those of trivalent lanthanide com-plexes.20 The selected bond lengths of the two complexestogether with those of [Gd(ODO3A)(H2O)]
+,21 and [Gd-(DO3A)]22 are listed in Table 2 for comparison. Note thatB3LYP calculations in combination with large-core ECPs areknown to overestimate Ln–N distances. For example, theaverage crystal and our DFT calculated (in parentheses) Eu–Nand Eu–O(carboxylate) bond distances for the EuDOTA−
complex are 2.68 Å (2.73 Å) and 2.38 Å (2.36 Å), respectively.An overestimation of the Eu–N bond distance by 0.05 Å and anunderestimation of the Eu–O bond distance by 0.02 Å comparedto crystal structural data have been observed in the present DFTstudy.
In the EuDO2A(H2O)3+ structure, the 3 inner-sphere coordi-
nated water molecules are arranged in a meridional configura-tion,3c,23 i.e. the 3 water molecules are on the same planeperpendicular to that of the basal N4 atoms. For the EuODO2A-(H2O)3
+ complex, the position of the middle coordinated watermolecule is slightly moved away from the plane perpendicular tothat of the basal N3O atoms. However, the Eu(III)–O (apicalH2O) bond distance is longer than the other two Eu(III)–O (sideH2O) bond distances for each complex. Thus, dinuclear species
formation could only result in the form of Ln2L2(μ-OH)2. Theformation of tri-hydroxo-bridged dinuclear species is unlikelyfor the Ln(III)-DO2A2− and Ln(III)-ODO2A2− complex systems.
Owing to the slow convergence of the geometry optimizationsfor the dinuclear Eu(III)-ODO2A2− systems in aqueous solutionas experienced by us and others,15f a Y(III) ion was used as asubstitute for the Eu(III) ion in these systems for DFT calcu-lations. The ionic radii of Y(III) (1.019 Å) and Eu(III) (1.066 Å)ions are similar,24 and they both belong to the rare-earth elementfamily due to chemical similarities, although the Y(III) ion hasno 4f electrons. Thus, the configurations of YDO2A+,YODO2A+, EuDO2A+ and EuODO2A+ are expected to besimilar, except that the Eu(III) complexes are 9-coordinated withthree inner-sphere coordinated H2O molecules, the Y(III) com-plexes are 8-coordinated with two inner-sphere coordinated H2Omolecules. Attempts to either add or reduce inner-sphere coordi-nated H2O molecules resulted in higher energies for both Eu(III)and Y(III) complexes. On the other hand, the lack of 4f electronsallows the DFT calculations of the Y(III) complexes to run fasterthan those for Ln(III) complexes.
The DFT calculated low-energy structures of the μ-OHbridged Y2(ODO2A)2(μ-OH) dinuclear species are shown inFig. 6. Two structures are illustrated for the mono-μ-OH bridgeddinuclear species: one with the lowest energy has one μ-OHbridge and a H2O–Y–Y–OH2 torsional angle of 82.21° (cis-[Y2(ODO2A)2(μ-OH)(H2O)2]
+, −2326.666586 a.u.; Fig. 6,upper-left). The other with higher energy has one μ-OH bridgeand a H2O–Y–Y–OH2 torsional angle of 148.54° (trans-
Fig. 3 Speciation diagrams of the La(III)-ODO2A2− (left), Eu(III)-ODO2A2− (middle) and Yb(III)-ODO2A2− (right) complex systems. [La(III)] =[Eu(III)] = [Yb(III)] = [ODO2A2−(L2−)] = 0.001 M.
upper-right). A structure of a di-μ-OH bridged dinuclear speciesis also shown in Fig. 6 (lower-middle, [Y2(ODO2A)2(μ-OH)2]).It is observed that each Y(III) ion is 8-coordinated in the threecomplex structures shown in Fig. 6. The first and the secondinclude 6-coordination by the ligand ODO2A2−, one by thebridged μ-OH ion and one by a water molecule. The thirdincludes 6-coordination by the ligand ODO2A2− and two by thebridged μ-OH ions.
The two inner-sphere coordinated water molecules in the[Y2(ODO2A)2(μ-OH)(H2O)2]
+ dinuclear species are in a stag-gered conformation with a torsional angle of 82.21° or 148.54°.It is possible that the structure with one additional inner-spherecoordinated water molecule on each Y(III) ion at the availablecoordinating space of the [Y(ODO2A)]2(μ-OH)2 dinuclearspecies is also present because both β22-3 and β22-4 values couldbe fitted from the Y(III)-ODO2A2− titration data. Selectedaverage bond distances and angles of these four complexes arelisted in Table S2 (ESI†).
Luminescence studies
Fig. S4 (ESI†) show the emission spectra of the EuODO2A+
complex at pH 6.1 in H2O and D2O, respectively, at room
temperature (λex = 395 nm). These spectra are very similar tothose measured at pH 7.0 and pH 5.0. The lifetime data for theemission peak at 615 nm of the EuODO2A+ complex in both
Fig. 5 DFT calculated EuDO2A(H2O)3+ (left two structures, top- and side-views) and EuODO2A(H2O)3
+ (right two structures, top- and side-views)structures.
Table 2 DFT calculated selected bond lengths (Å) in [EuDO2A(H2O)3]+ and [EuODO2A(H2O)3]
+, as compared with those of [GdODO3A(H2O)]and [GdDO3A]3Na2CO3 moleculesa,b,c
a For DO2A2− and ODO2A2−, Oa is the apical H2O molecule. O1 and O3 are the two side H2O molecules. O2 and O4 are the oxygen atoms of thetwo COO− groups. For DO3A3− (1,4,7,10-tetraazacyclododecane-1,4,7-triacetate ion) and ODO3A3− (also abbreviated as DOTRA, 1-oxa-4,7,10-triazacyclododecane-4,7,10-triacetate ion), O3 is the oxygen atom of the middle COO− group. b [Gd(ODO3A)] data were obtained from crystal data,ref. 21. ODO3A3− is 7-coordinated, the complex has one apical coordinated water molecule (Oa), and the remaining coordination site is shared with acarboxylate group of a neighboring complex (O1). cValues are from reference crystallographic data, ref. 22. The bond distances are the two Eu–Obonds with the coordinated carbonate anions.
Fig. 6 DFT calculated structures of cis-[Y2(ODO2A)2(μ-OH)(H2O)2]+
H2O and D2O solutions at pH 7.0 could be fitted to a singleexponential luminescence decay kinetic equation. Similarly, theemission spectra (λex = 369 nm, Fig. S5, ESI†) and the emissionlifetime data at 544 nm of the TbODO2A+ complex at pH 7.0 inH2O and D2O, respectively, have been obtained at room tempera-ture. Table 3 lists the lifetime data and the numbers of inner-sphere coordinated water molecules (q) of the EuODO2A-(H2O)q
+ and the TbODO2A(H2O)q+ complexes obtained using
the lifetime data by employing different previously establishedempirical equations.25 For comparison purposes, those of theEuDO2A(H2O)q
+ complex are also included.11
The q values obtained for the EuODO2A(H2O)q+ and
EuDO2A(H2O)q+ complexes from empirical equations qa–d were
all greater than that from equation qe. The early empiricalequation qa by Sudnick and Horrocks quoted an uncertaintyvalue of ±0.5 for the q values obtained probably due to thelower number of experimental data used.25a This has beenrefined by equation qe with a standard error of ±0.1 inq-values.25e For the derivation of the parameters for equation qb
by Beeby et al.,25b no species with q values greater than 6 wereemployed and the results for q values tend to be on the higherside. For equations qc and qd, it assumes that the quenchingeffects in D2O are similar for all Eu3+ complexes and certainlyprovided only rough estimates of the q values.25c,d It is notedthat these five equations all assumed that the contributions fromthe alcoholic O–H, amine N–H, alkane C–H and amide carbonyloscillators in the first coordination sphere of the Eu3+ complexwere negligible which is certainly not always true.11,25f,g
Ideally, lifetime measurements should be carried out withoutdeuterium exchange of NH groups, so that these oscillators wouldcontribute to the same extent to quench lanthanide luminescence.It has been found that deuterium exchange is very slow for NHgroups coordinated to lanthanide ions.26 To check the effect of theN–H oscillator, the lifetime of EuODO2A+ in freshly preparedD2O solution without drying and redissolving was measured andit was found that τD2O (1.42 ms) was the same as that with dryingand redissolving. This indicates that H/D exchange in theEuODO2A+ complex at pH 7.0 is probably slow within the D2Osolution preparation time period and the lifetime result obtained inD2O solution is valid for the estimation of q value.
The averaged q value for the EuODO2A(H2O)q+ complex is
4.1 which is greater than that of the EuDO2A(H2O)q+ complex
(q = 3.0), indicating that the complex has structures with anaverage of four inner-sphere coordinated water molecules. Thus,although the DFT calculated structure has the lowest energy withq = 3 (vide supra), it might still be possible that this complex
could have equilibrium structures accommodating four or moreinner-sphere coordinating water molecules. This might be due tothe fact that the 12-membered macrocyclic ring of ODO2A2− issmaller, less symmetrical, not pre-organized20e and more rigidwhich may lead to more coordination space available for moreinner-sphere coordinated water molecules on the Ln(III) ion.Similar results have been observed for other Eu(III) complexes ofmultidentate ligands with oxaaza backbones.14,27,28 Note that theDFT calculated EuODO2A+ structure showed that the Eu–O5bond distance (2.55 Å) is shorter than the three Eu–N(6–8) bonddistances (averaged 2.72 Å) leading to a less-symmetricalcoordination environment as compared to that of the EuDO2A+
complex. [Care should be exercised here to note that an overesti-mation of the Eu–N (e.g. 0.05 Å) bond distance and an under-estimation of the Eu–O bond distance (e.g. 0.02 Å) are known inthe present DFT study (vide supra).] This could allow othercomplex conformations with similar or slightly higher energiesto occur and with one more coordinated water molecule. Inaddition, it is also possible that the oxygen atom and the second-ary nitrogen atom in the macrocycle may have additional unusualeffects in helping the complex to relax from the excited state.
For comparison purposes, the number of inner-sphere coordi-nated water molecules of TbODO2A+ has also been measured25a,b,d
to be 2.9 which is consistent with the expected q = 3 and previouslyreported values of several Tb(III) complexes of linear and macrocyc-lic multidentate ligands.28 For example, the average q value in thepH range 4–11 is 1.8 ± 0.2 for TbDO3A (DO3A3− is 1,4,7,10-tetra-azacyclododecane-1,4,7-triacetate ion) because DO3A3− is a7-coordinating ligand and Tb(III) ion is 9-coordinated.28b
An interesting pH-dependence effect is observed for the lumine-scence emission spectra of the EuODO2A+ complex, i.e. the spec-tral band intensity at 615 nm increases with increasing pH (Fig. 7).The lifetimes (in ms) are roughly constant from pH 5 to pH 7 andincrease with increasing pH afterwards: 0.218 (pH 4.0), 0.220(pH 5.0), 0.220 (pH 6.1), 0.219 (pH 7.0), 0.232 (pH 8.0), 0.257(pH 9.0), 0.283 (pH 10.0), and 0.291 (pH 11.0). The gradualincrease with increasing pH of lifetimes after pH 7.0 is consistentwith the fact that the number of O–H oscillators decreases withincreasing pH due to μ-OH bridged dinuclear species as well asmono-coordinated-OH formations. Similar observations have beenfound for the TbODO2A+ complex (data not shown).
Conclusions
The log Kf values of the LnODO2A+ complexes increaseroughly with increasing lanthanide atomic number which
Table 3 The luminescence lifetimes (τ, ms) and the numbers of inner-sphere coordinated water molecules (q) of the EuODO2A(H2O)q+ complex
indicates that the bonding is mainly electrostatic in nature withsome macrocycle size selectivity. The LnODO2A+ complex for-mations are relatively faster as compared to those of theLnDO2A+ complexes. This allows the determinations of for-mation constants of various hydrolytic species according to anunprecedented general hydrolysis scheme involving the stepwiseand simultaneous formations of mono- and di-nuclear macrocyc-lic LnODO2A+ complexes. The dinuclear structures are gener-ally more abundant than the mononuclear ones. Massspectrometry confirmed the presence of [Eu(ODO2A)]+, [Eu-(ODO2A)(OH) + H]+, [Eu2(ODO2A)2(OH2)2 + H]+, [Eu-(ODO2A)(OH)2]
− and [Eu2(ODO2A)2(OH2)3]− species at pH
higher than 7. The density function theory (DFT) calculatedstructures of the EuODO2A(H2O)3
+ and EuDO2A(H2O)3+ com-
plexes indicate that the 3 inner-sphere coordinated water mo-lecules are arranged in a meridional configuration, i.e. the 3 watermolecules are nearly on the same plane perpendicular to that ofthe basal N3O or N4 atoms. However, luminescence lifetimestudies reveal that the EuODO2A+ and TbODO2A+ complexeshas 4.1 and 2.9 inner-sphere coordinated water molecules,respectively, indicating that other equilibrium species are alsopresent for the EuODO2A+ complex. Additional DFT calcu-lations show that each Y(III) ion is 8-coordinated in the threepossible cis-[Y2(ODO2A)2(μ-OH)(H2O)2]
+, trans-[Y2(ODO2A)2-(μ-OH)(H2O)2]
+ and [Y2(ODO2A)2(μ-OH)2] dinuclear complexstructures. The first and second include 6-coordination by theligand ODO2A, one by the bridged μ-OH ion and one by awater molecule. The third includes 6-coordination by the ligandODO2A2− and two by the bridged μ-OH ions. The two inner-sphere coordinated water molecules in the cis- and trans-[Y(ODO2A)]2(μ-O(H2O)2]
+ dinuclear species are in a staggeredconformation with torsional angles of 82.21° and 148.54°,respectively. These results will be very useful for the design ofligands that form dinuclear Ln(III) complexes for various analyti-cal and biomedical applications.
Experimental
Materials and standard solutions
Analytical reagent-grade chemicals and buffers, unless otherwisestated, were purchased from Sigma (St. Louis, MO, USA),
Aldrich (Milwaukee, WI, USA) or Merck (Dammstadt,Germany) and were used as received without further purification.Disodium ethylenediaminetetraacetic acid (Na2H2EDTA) waspurchased from Fisher. The ligand H2ODO2Awas prepared andpurified as described below. Carbonate-free deionized water wasused for all solution preparations.
The concentration of the H2ODO2A stock solution (ca.0.01 M) was determined by pH titration using a standard tetra-methylammonium hydroxide solution (0.1 M), and was alsochecked by complexometric back-titration.11 The concentrationsof the lanthanide nitrate stock solutions were ca. 0.01 M andwere standardized by EDTA titration using xylenol orange as theindicator. The EDTA solution was standardized by titrating acalcium carbonate primary standard solution (first dissolved inHCl solution) at pH 10 using calmagite as the indicator.
The 0.1 M tetramethylammonium hydroxide solution was pre-pared by diluting a 20% (CH3)4NOH–methanol solutionobtained from Aldrich (carbonate-free). The aqueous (CH3)4-NOH solution was standardized by using reagent grade primarystandard potassium hydrogen phthalate. A 0.1 M HCl solutionwas prepared by diluting a reagent grade HCl solution and stan-dardized by using the standard (CH3)4NOH solution. A 1.0 Mstock solution of tetramethylammonium chloride (Aldrich) wasprepared and diluted to 0.1 M for each titration to maintain aconstant ionic strength (0.1 M).
Synthesis of H2ODO2A
The compound was synthesized according to a publishedmethod with minor modification.15 N,N,N-Tri(p-toluensulfonyl)-diethylenetriamine was first prepared and purified using the tri-tosylated diethylenetriamine (yield 75%, m.p. 176–177 °C) andwas added equivalent amount of 1,5-bis(p-toluenesulfonyl)-3-oxapentane (yield 92%, m.p. 81–82 °C) to form the macro-cycle 4,7,10-tris(p-tolysulphonyl)-1-oxa-4,7,10-triazacyclododecane(yield 77%, m.p. 196–198 °C). 1-Oxa-4,7,10-triazacyclodo-decane was then obtained by reacting 4,7,10-tris(p-tolysulpho-nyl)-1-oxa-4,7,10-triazacyclo-dodecane with sulfuric acid (yield72%, m.p. 78–79 °C). 4,10-Dicarboxymethyl-1-oxa-4,7,10-tri-azacyclododecane (ODO2A) in the hydrochloride form wasfinally obtained by carboxymethylation of 1-oxa-4,7,10-triaza-cyclo-dodecane with bromoacetate and recrystallization inethanolic HCl solution (yield 35%). Anal. Calc. for H2ODO2A·2.7HCl·H2O (C12H23N3O5·2.7HCl·H2O): C, 35.51; H, 6.88;N, 10.36. Found: C, 35.53; H, 6.89; N, 10.12. NMR(D2O-DSS),
Acid–base and complex formation titrations were carried out at aconstant ionic strength of 0.10 M (CH3)4NCl using reported pro-cedures.11 A model 720 Metrohm Titroprocessor in conjunctionwith Metrohm Combination Electrode was employed to monitorthe pH. Generally, equal molar concentrations of the ligand andthe metal ion were employed (∼1.0 mM) for the titration. Theionic strength of the solution was adjusted to 0.1 M using 1 M
Fig. 7 Luminescence spectra of the EuODO2A+ complex at varioussolution pH. λex = 395 nm.
(CH3)4NCl. The (CH3)4NOH solution was delivered from a10 mL automatic Brinkmann Metrohm Model 665 Dosimatburet with a reading accuracy of ±0.001 mL. Most of the titrationreactions reached equilibrium rather quickly during the potentio-metric titration process, and the titrations were completed within18 000–40 000 s.
The pH-metric titration data were used to calculate the step-wise ligand protonation constants defined in eqn (1), where n =1–4 for ODO2A (L):
Kn ¼ ½HnL�=ð½Hn�1L�½Hþ�Þ ð1ÞAll equilibrium calculations were performed using the
program Hyperquad2008 Version 5.2.15. The averaged valuesare presented together with the standard deviations calculatedfrom valid data points. Speciation diagrams were generatedusing Hyss 2006 software.
Mass spectral measurements
Mass spectra of the Eu(III)-ODO2A2− complex system at variouspH were acquired by direct infusion (10 μL). ESI(+)MS experi-ments were carried out using a LTQ-Orbitrap hybrid tandemmass spectrometer (ThermoFisher, USA) equipped with an elec-trospray ionization (ESI) source operating in positive ion mode.The parameters of ESI(+) included 4.0 kV for ion spray voltage,200 °C for capillary temperature, and 3–5 erb for sheath gasflow rate. The mass spectra were collected over the mass rangeof m/z 150–2000 at a resolving power of 30 000. The collecteddata were analyzed using Xcalibur software (ThermoFisher,USA). ESI(−)-MS measurements were performed using a micro-mass triple quadrupole mass spectrometer equipped with an ESIsource operating in negative ion mode. The conditions used forthe ESI(−) interface included 500 (L h−1) for desolvation gas,52 (L h−1) for cone gas, 3.2 kV for capillary voltage, 30 V forcone voltage and 250 °C for desolvation temperature.
NMR determinations of protonation sites
Solutions of ligand (0.1 mM) for 1H NMR titrations were pre-pared in D2O solution, and pD was adjusted with DCl or carbon-ate-free NaOD. The apparent pD (pH) of the ligand solutionswas determined with a microelectrode and the final pD valuewas obtained from the equation pD = pH + 0.40. All chemicalshifts were referenced to the 3-(trimethylsilyl)propionic acid-d4sodium salt.
Luminescence measurements
Steady-state luminescence experiments were carried out on anEdinburgh Instruments FSP920 fluorescence system equippedwith a 450 W xenon arc lamp as the illumination source. Emis-sion light was collected into a TMS300 Czerny-Turner configur-ation double grating monochromator and detected by aHamamatsu R928P photomultiplier tube in the visible wave-length range. Spectra were recorded by use of F900 Fluorescencespectrometer software. Luminescence lifetimes were recorded byuse of a μF920H flashlamp (lamp frequency: 100 Hz) as theexcitation source with the multiple-channel single photon
counting mode (MCS). Data was fitted by a nonlinear least-squares iterative technique (Marquardt-Levenberg algorithm).
For the LnODO2A+ systems (Ln = Eu or Tb), all solutionswere prepared from stock solutions of Ln(NO3)3 (∼0.01 M) andH2ODO2A (∼0.01 M). The molar ratio of the H2ODO2A andLn3+ was ca. 1.00 : 0.99 to make sure that the ligand was inslight excess. All solutions were allowed to equilibrate for atleast 12 h to ensure complete formation reaction prior to lumine-scence measurements. The sample solutions in D2O were pre-pared by first evaporating the aqueous solutions in appropriateflasks to dryness using a high vacuum system. To each flask,D2O was added to dissolve the solids, the solution was equili-brated for at least 1 h, dried again, and finally dissolved in D2Oto a cuvette for luminescence measurements. Freshly preparedcomplex solutions in D2O without drying and redissolving werealso prepared for comparative measurements. The correspondingnumbers of inner-sphere coordinated water molecules werecalculated using the lifetime data in H2O and D2O with esta-blished empirical equations.19 All data were checked at leasttwice carefully.
Density function theory (DFT) computations of lanthanidecomplex structures
All calculations were performed employing HF and hybrid DFTwith the B3LYP exchange correlation functional and theGaussian 09 package. Full geometry optimizations of the [Eu-(DO2A)(H2O)3]
+, [Eu(ODO2A)(H2O)3]+, cis-[Y2(ODO2A)2-
(μ-OH)(H2O)2]+, trans-[Y2(ODO2A)2(μ-OH)(H2O)2]
+ and[Y2(ODO2A)2(μ-OH)2] systems were obtained in vacuum byusing the 3-21G and 6-31G* basis sets for carbon, hydrogen,nitrogen, and oxygen. For these complexes, for example, thelowest energy crystallographic structure of EuDOTA− [squareantiprismatic, SAP, Λ(δδδδ)] from Cambridge Structural Data-base (CSD) was used to construct the initial guessed structuresof the EuDO2A+ complex by removing two opposite carboxylategroups. The initial guessed structures were refined by a lowerlevel HF/3-21G method and then refined by DFT calculationsusing the B3LYP/6-31G* method to find the lowest energycomplex structures. The quasi-relativistic effective core potential(ECP) of Stuttgart RSC 1997 ECP and the related [5s4p3d]-GTO valence basis set were applied to europium and yttriumatoms. This ECP treats [Kr]4d104fn as fixed core, while only the5s5p6s5d6p shell is taken into account explicitly. The procedureinvolved initial HF/3-21G optimization of the molecular systemand refinement by using B3LYP/6-31G* basis set, both in gasphase. The geometries of the complexes were then fully opti-mized by B3LYP/6-31G* and ECP in aqueous solutions byusing a polarizable continuum model (PCM). Frequency calcu-lations with zero-point corrections were also performed with thelowest-energy conformers of the LnL+ (Ln = Eu, Y; L = DO2A2−,ODO2A2−) complexes.
Acknowledgements
The authors wish to thank the National Science Council and theAtomic Energy Council of the Republic of China (Taiwan) forfinancial support (grant numbers NSC-98-2113-M-010-001-
MY3, NSC 96-NU-7-009-003, and NSC-100-2811-M-010-003)of this work.
References
1 (a) P. Caravan, J. J. Ellison, T. J. McMurry and R. B. Lauffer, Chem.Rev., 1999, 99, 2293–2352; (b) C. A. Chang, Eur. J. Solid State Inorg.Chem., 1991, 28, 237–244; (c) C. A. Chang, L. Francesconi,M. F. Malley, K. Kumar, J. Z. Gougoutas, M. F. Tweedle, D. W. Lee andL. J. Wilson, Inorg. Chem., 1993, 32, 3501–3508; (d) C. A. Chang,P. Sieving, A. D. Watson, T. M. Dewey, T. B. Karpishin andK. N. Raymond, J. Magn. Reson. Imaging, 1992, 2, 95–98;(e) C. A. Chang, Invest. Radiol., 1993, 28, S21–S27; (f ) J. Varadarajan,S. Crofts, J. D. Fellmann, J. C. Carvalho, S. Kim, C. A. Chang andA. Watson, Invest. Radiol., 1994, 29, S18–S20; (g) T. Z. Lee,T. H. Cheng, M.-H. Ou, C. A. Chang, G. C. Liu and Y. M. Wang, Magn.Reson. Chem., 2004, 42, 329–336; (h) V. Jacques and J. F. Desreux, Top.Curr. Chem., 2002, 221, 123–164; (i) P. Hermann, J. Kotek, V. Kubicekand I. Lukes, Dalton Trans., 2008, 3027–3047.
2 (a) J.-C. G. Bunzli, Chem. Rev., 2010, 110, 2729–2755;(b) C. P. Montgomery, B. S. Murray, E. J. New, R. PAL and D. Parker,Acc. Chem. Res., 2009, 42, 925–937; (c) C. M. G. Dos Santos,A. J. Harte, S. J. Quinn and T. Gunnlaugsson, Coord. Chem. Rev., 2008,252, 2512; (d) L. E. Jennings and N. J. Long, Chem. Commun., 2009,3511–3524; (e) S. Mizukami, K. Tonai, M. Kaneko and K. Kikuchi,J. Am. Chem. Soc., 2008, 130, 14376–14377.
3 (a) P. D. Beer and P. A. Gale, Angew. Chem., Int. Ed., 2001, 40, 486;(b) T. L. Esplin, M. L. Cable, H. B. Gray and A. Ponce, Inorg. Chem.,2010, 49, 4643–4647; (c) M. L. Cable, J. P. Kirby, K. Sorasaenee,H. B. Gray and A. Ponce, J. Am. Chem. Soc., 2007, 129, 1474–1475;(d) R. Pal, D. Parker and L. C. Costello, Org. Biomol. Chem., 2009, 7,1525–1528.
4 (a) J. R. Morrow, Comments Inorg. Chem., 2008, 29, 169;(b) A. K. Yatsimirsky, Coord. Chem. Rev., 2005, 249, 1997;(c) B. N. Trawick, A. T. Daniher and J. K. Bashkin, Chem. Rev., 1998,98, 939; (d) N. H. Williams, B. Takasaki, M. Well and M. J. Chin, Acc.Chem. Res., 1999, 32, 485; (e) A. Roigk, R. Hettich andH. J. Schneider, Inorg. Chem., 1998, 37, 751; (f ) P. Hurst,B. K. Takasaki and J. Chin, J. Am. Chem. Soc., 1996, 118, 9982;(g) M. Komiyama and T. Takarada, Acid-promoted peptide bondhydrolysis, in Metal Ions in Biological Systems, ed. A. Sigel andH. Sigel, Marcel Dekker, New York, 2003, vol. 40, ch. 10, p. 355;(h) H.-J. Schneider and A. K. Yatsimirski, Lanthanide-catalyzedhydrolysis of phosphate esters and nucleic acids, in Metal Ions in Bio-logical Systems, ed. A. Sigel and H. Sigel, Marcel Dekker, New York,2003, vol. 40, ch. 11, p. 369; (i) M. Komiyama, Sequence-selectivescission of DNA and RNA by lanthanide ions and their complexes, inMetal Ions in Biological Systems, ed. A. Sigel and H. Sigel, MarcelDekker, New York, 2003, vol. 40, ch. 12, p. 463.
5 C. F. Baes and R. E. Mesmer, The Hydrolysis of Cations, Wiley,New York, 1976, pp. 129–146.
6 Z. Zhang, Chem. Commun., 2001, 2521–2529.7 G. D. Klungness and R. H. Byrne, Polyhedron, 2000, 19, 99–107.8 (a) R. V. Southwood-Jones and A. E. Merbach, Inorg. Chim. Acta, 1978,30, 77–82; (b) C. A. Chang, Y.-P. Chen and C.-H. Hsiao, Eur. J. Inorg.Chem., 2009, 1036–1042.
9 (a) K. New, C. M. Andolina and J. R. Morrow, J. Am. Chem. Soc.,2008, 130, 14861–14871; (b) F. Aguilar-Prez, P. Gmez-Tagle,E. Collado-Fregoso and A. K. Yatsimirsky, Inorg. Chem., 2006, 45,9502–9517.
10 (a) C. A. Chang, B. H. Wu and B. Y. Kuan, Inorg. Chem., 2005, 44,6646–6654; (b) C. A. Chang, B. H. Wu and C.-H. Hsiao, Eur. J. Inorg.Chem., 2009, 1339–1346.
11 C. A. Chang, F.-K. Shieh, Y.-L. Liu, Y.-H. Chen, H.-Y. Chen andC.-Y. Chen, J. Chem. Soc., Dalton Trans., 1998, 3243–3248.
12 (a) C. A. Chang and M. E. Rowland, Inorg. Chem., 1983, 22, 3866–3869; (b) C. A. Chang and V. C. Ochaya, Inorg. Chem., 1986, 25, 355–358; (c) C. A. Chang, P. H. L. Chang and S. Qin, Inorg. Chem., 1988,27, 944–946; (d) C. A. Chang, V. O. Ochaya and V. C. Sekhar, J. Chem.Soc., Chem. Commun., 1985, 1724–1725.
13 M. T. S. Amorim, J. R. Ascenso, R. Delgado and J. J. R. Fraústo daSilva, J. Chem. Soc., Dalton Trans., 1990, 3449–3455.
14 R. C. Holtz, S. L. Klakamp, C. A. Chang and W. D. Horrocks Jr., Inorg.Chem., 1990, 29, 2651–2658.
15 (a) M. Purgel, Z. Baranyai, A. de Blas, T. Rodrıguez-Blas, I. Banyai,C. Platas-Iglesias and I. Toth, Inorg. Chem., 2010, 49, 4370–4382;(b) U. Cosentino, G. Moro, D. Pitea, A. Villa, P. Carlo Fantucci,A. Maiocchi and F. Uggeri, J. Phys. Chem. A, 1998, 102, 4606–4614;(c) O. V. Yazyev, L. Helm, V. G. Malkin and O. L. Malkina, J. Phys.Chem. A, 2005, 109, 10997–11005; (d) L. Smentek and B. A. Hess Jr.,Collect. Czech. Chem. Commun., 2008, 73, 1437–1456; (e) J. Notni,K. Pohle, J. A. Peters, H. Görls and C. Platas-Iglesias, Inorg. Chem.,2009, 48, 3257–3267; (f ) M. Mato-Iglesias, T. Rodrıguez-Blas, C. Platas-Iglesias, M. Starck, P. Kadjane, R. Ziessel and L. Charbonniere, Inorg.Chem., 2009, 48, 1507–1518.
16 (a) M. T. S. Amorim, S. Chaves, R. Delgado and J. Dasilva, J. Chem.Soc., Dalton Trans., 1991, 3065–3072; (b) S. Chaves, A. Cerva andR. Delgado, J. Chem. Soc., Dalton Trans., 1997, 4181–4189.
17 M. Kodama, T. Koike, A. B. Mahatma and E. Kimura, Inorg. Chem.,1991, 30, 1270–1273.
18 A. Roca-Sabio, M. Mato-Iglesias, D. Esteban-Gómez, E. Toth, A. deBlas, C. Platas-Iglesias and T. Rodrıguez-Blas, J. Am. Chem. Soc., 2009,131, 3331–3341.
19 (a) C.-C. Lin, C.-L. Chen, K.-Y. Liu and C. A. Chang, Dalton Trans.,2011, 40, 6268–6277; (b) V. C. Sekhar and C. A. Chang, Inorg. Chem.,1986, 25, 2061–2065.
20 (a) L. J. Charbonnière, R. Schurhammer, S. Mameri, G. Wipff andR. F. Ziesse, Inorg. Chem., 2005, 44, 7151–7160; (b) U. Cosentino,A. Villa, D. Pitea, G. Moro, V. Barone and A. Maiocchi, J. Am. Chem.Soc., 2002, 124, 4901–4909; (c) D. J. Feller, J. Comp. Chem., 1996, 17,1571–1586; (d) K. L. Schuchardt, B. T. Didier, T. Elsethagen, L. Sun,V. Gurumoorthi, J. Chase, J. Li and T. L. Windus, J. Chem. Inf. Model.,2007, 47, 1045–1052; (e) H. Y. Lee, C.-L. Chen and C. A. Chang,Dalton Trans., submitted for publication.
21 M. R. Spirlet, J. Rebizant, X. Wang, T. Jin, D. Gilsoul, R. N. Muller andJ. F. Desreux, J. Chem. Soc., Dalton Trans., 1997, 497–500.
22 C. A. Chang, L. Francesconi, M. F. Malley, K. Kumar, J. Z. Gougoutas,M. F. Tweedle, D. W. Lee and L. J. Wilson, Inorg. Chem., 1993, 32,3501–3508.
23 T. Le Borgne, P. Altmann, N. Andre, J.-C. G. Bunzli, G. Bernardinelli,P.-Y. Morgantini, J. Weber and C. Piguet, J. Chem. Soc., Dalton Trans.,2004, 723.
24 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor.Gen. Crystallogr., 1976, 32, 751–767.
25 (a) W. DeW. Horrocks Jr. and D. R. Sudnick, Acc. Chem. Res., 1981, 14,384–392; (b) A. Beeby, I. M. Clarksin and R. S. Dickins, et al., J. Chem.Soc., Perkin Trans. 2, 1999, 493–503; (c) P. P. Barthelemy andG. R. Choppin, Inorg. Chem., 1989, 28, 3354–3357; (d) T. Kimura andY. Kato, J. Alloys Compd., 1995, 225, 284–287; (e) R. M. Supkowskiand W. DeW. Horrocks Jr., Inorg. Chim. Acta, 2002, 340, 44–48;(f ) D. Parker and J. A. G. Williams, J. Chem. Soc., Dalton Trans., 1996,3613; (g) R. S. Dickins, D. Parker, A. S. de Sousa and J. A. G. Williams,Chem. Commun., 1996, 697.
26 A. Rodríguez-Rodríguez, D. Esteban-Gómez, A. de Blas, T. Rodríguez-Blas, M. Fekete, M. Botta, R. Tripier and C. Platas-Iglesias, Inorg.Chem., 2012, 51, 2509–2521.
27 Y. Wang and W. DeW. Horrocks Jr., Inorg. Chim. Acta, 1997, 263,309–314.
28 (a) C. A. Chang, H. G. Brittain, J. Telser and M. F. Tweedle, Inorg.Chem., 1990, 29, 4468–4473; (b) X. Zhang, C. A. Chang, H. G. Brittain,J. M. Garrison, J. Telser and M. F. Tweedle, Inorg. Chem., 1992, 31,5597–5600.
